Space Weather Effects on the Earth’s Upper Atmosphere: Short Report on Ionospheric Storm Effects at Middle Latitudes

: During geomagnetic storm events, the highly variable solar wind energy input in the magnetosphere signiﬁcantly alters the structure of the Earth’s upper atmosphere through the interaction of the ionospheric plasma with atmospheric neutrals. A key element of the ionospheric storm-time response is considered to be the large-scale increases and decreases in the peak electron density that are observed globally to formulate the so-called positive and negative ionospheric storms, respectively. Mainly due to their signiﬁcant impact on the reliable performance of technological systems, ionospheric storms have been extensively studied in recent decades, and cumulated knowledge and experience have been assigned to their understanding. Nevertheless, ionospheric storms constitute an important link in the complex chain of solar-terrestrial relations. In this respect, any new challenge introduced in the ﬁeld by a better understanding of the geospace environment, new modeling and monitoring capabilities and/or new technologies and requirements also introduces new challenges for the interpretation of ionospheric storms. This paper attempts a brief survey of present knowledge on the fundamental aspects of large-scale ionospheric storm time response at middle latitudes. Further attention is paid to the results obtained regarding the critical role that solar wind conditions which trigger disturbances may play on the morphology and the occurrence of ionospheric storm effects.


Introduction
The Earth's ionosphere is the ionized part of the upper atmosphere, extending from about 60 km to 1000 km altitude imbedded in the neutral atmosphere. Ionospheric ionization is produced and further stratified through solar radiation, with the maximum concentration of the ionization to occur in the ionospheric F region. As the inner edge of the magnetosphere with clear links to the neutral atmosphere, the ionosphere plays a critical role in solar-terrestrial relationships, merging the influence from both above and below. In this respect, the transient changes imposed by space weather events in the near-Earth space environment lie among most challenging topics for both scientific and operational applications, e.g., [1][2][3].
During disturbed space weather conditions, the highly variable solar wind energy input in the magnetosphere significantly impacts the upper atmosphere's state through the interaction of the ionospheric plasma with atmospheric neutrals. Most dramatic changes follow the occurrence of geomagnetic storms, during which, a variety of photo-chemical and chemical reactions, as well as dynamical and electrodynamical processes that are activated in the magnetosphere-ionosphere-thermosphere system, drive the exchange and transportation of mass, momentum, and energy between the system's elements, and consequently, substantially alter the ionospheric structure. A key element of the ionospheric response to geomagnetic storm events comes in the form of disturbances in the peak electron density (NmF2) and column density (i.e., the total electron content-TEC): largescale increases and decreases in the two parameters are observed globally to formulate the so-called positive and negative ionospheric storms, respectively.

Materials and Methods
This work is greatly based on bibliographic research. Complementary results are obtained through the analysis of two ionospheric characteristics; the peak electron density in terms of the critical frequency foF2, and the height of the peak electron density hmF2 obtained by a network of European stations listed in Table 1. foF2 is related to the F-peak electron density NmF2 by NmF2/m −3 = 1.24 × 10 10 × (foF2/MHz) 2 . The analyzed period includes the years 2008-2020 that fully cover the evolution of solar cycle 24 (see also Figure 1). Ionospheric disturbances are justified through the comparison between the observed values with their monthly medians. Disturbances are quantified by the following metric: DfoF2% = [(foF2obs − foF2median)/foF2median] × 100 for foF2. Ionospheric observations were retrieved from the Global Ionosphere Radio Observatory (GIRO-http://giro.uml.edu/, accessed on 30 December 2021) data repository [5], particularly the FastChart database. The level of the geomagnetic activity is determined by records of Dst index available from the World Data Center (WDC) for Geomagnetism, Kyoto (http://wdc.kugi.kyoto-u.ac.jp/dstdir/, accessed on 30 December 2021). Geomagnetic storms may be grouped with respect to their intensity, as in Table 2.

Results
Geomagnetic storms are initiated by the arrival of geo-effective solar wind flows at the Earth's vicinity (see, for instance [6][7][8][9][10][11][12][13][14][15]). Geo-effectiveness is determined by a combination of solar wind velocity and southward orientation of the Interplanetary Magnetic Field (IMF), with the second to hold the most important role because of its far greater variability. Magnetic reconnection is the key physical process responsible for transferring solar wind energy into the magnetosphere during geomagnetic storms, being strongest under southward IMF orientation: southward directed IMF interconnects with the Earth's magnetopause northward magnetic field and solar wind drags the interconnected magnetic fields and plasma downstream (i.e., in the antisunward direction); the open magnetic fields then reconnect in the magnetotail. The relative rate of the reconnection in the magnetopause and magnetotail sets off a global convection cycle of magnetic flux between the dayside and nightside magnetosphere. The convection cycle (also known as Dungey cycle [16]) and related disturbances induce plasma instabilities, wave growth and wave-particle interactions, also giving rise to a wide range of global processes, including the buildup of storm-time ring current, the activation of the field-aligned current system and the auroral electrojects, the variability of radiation belt intensities, and enhanced particle precipitation into lower latitudes. Through such processes, the magnetospheric disturbances are projected down to the ionosphere. The stronger the southward IMF component and the stronger the solar wind velocity convecting the magnetic field, the more strongly the solar-wind-magnetosphere-ionosphere system is driven. This dependence is reflected in the magnetospheric energy input functions, such as the epsilon parameter of Akasofu, the most widely used one, according to which the solar wind energy input is determined by the solar wind speed, the IMF magnitude, and the so-called clock angle of the IMF orientation perpendicular to the Sun-Earth line [17][18][19]. The solar wind energy entering the magnetosphere is then dissipated via a variety of dynamic processes in the plasma sheet,

Results
Geomagnetic storms are initiated by the arrival of geo-effective solar wind flows at the Earth's vicinity (see, for instance [6][7][8][9][10][11][12][13][14][15]). Geo-effectiveness is determined by a combination of solar wind velocity and southward orientation of the Interplanetary Magnetic Field (IMF), with the second to hold the most important role because of its far greater variability. Magnetic reconnection is the key physical process responsible for transferring solar wind energy into the magnetosphere during geomagnetic storms, being strongest under southward IMF orientation: southward directed IMF interconnects with the Earth's magnetopause northward magnetic field and solar wind drags the interconnected magnetic fields and plasma downstream (i.e., in the antisunward direction); the open magnetic fields then reconnect in the magnetotail. The relative rate of the reconnection in the magnetopause and magnetotail sets off a global convection cycle of magnetic flux between the dayside and nightside magnetosphere. The convection cycle (also known as Dungey cycle [16]) and related disturbances induce plasma instabilities, wave growth and wave-particle interactions, also giving rise to a wide range of global processes, including the build-up of storm-time ring current, the activation of the field-aligned current system and the auroral electrojects, the variability of radiation belt intensities, and enhanced particle precipitation into lower latitudes. Through such processes, the magnetospheric disturbances are projected down to the ionosphere. The stronger the southward IMF component and the stronger the solar wind velocity convecting the magnetic field, the more strongly the solarwind-magnetosphere-ionosphere system is driven. This dependence is reflected in the magnetospheric energy input functions, such as the epsilon parameter of Akasofu, the most widely used one, according to which the solar wind energy input is determined by the solar wind speed, the IMF magnitude, and the so-called clock angle of the IMF orientation perpendicular to the Sun-Earth line [17][18][19]. The solar wind energy entering the magnetosphere is then dissipated via a variety of dynamic processes in the plasma sheet, the ring current and the ionosphere, with the latter being the dominant sink of the solar wind energy input through storm currents and particle precipitation, both contributing to the heating of the upper atmosphere. This energy dissipation in the high-latitude ionosphere generates numerous upper atmospheric changes that affect all state parameters.
(ii) High-speed (~750-800 km/s) solar wind streams (HSSs) emanating from solar coronal holes and corotating interaction regions (CIRs). CIRs are associated with HSSs that interact with slow-speed (~300-400 km/s) streams and are characterized by embedded and amplified Alfvén waves. CIRs usually lead to weak or moderate geomagnetic storms, mainly in the declining phase of the solar cycle. Moreover, they tend to evolve gradually and recover in longer time scales than CME-driven events. It is suggested that the CIR storms are driven by the magnetic reconnection of the southward component of the interplanetary Alfvén waves to the Earth's dayside magnetopause fields [6][7][8][9][10][11][12][13][14][15].

Ionospheric Storms-Morphology and Background Mechanisms
The main drivers of the ionospheric disturbances during a geomagnetic storm can be roughly categorized by four different methods: (i) enhanced high-energy particle precipitation; (ii) enhanced ionospheric electric currents and resulting Joule heating; (iii) enhanced electric fields; (iv) frictional heating at high latitudes, primarily induced by enhanced magnetospheric convection. While the direct effects of these mechanisms are mainly important at high and maybe at low (see for instance electric fields) latitudes, it has been argued that Joule and particle heating is globally the most critical factor causing the F region effects of geomagnetic storms [20][21][22][23]. Such effects are still activated in high latitudes, but they are transported to lower latitudes through wind circulation and waves [21].
Long-lasting negative storm effects are the most typical feature of the ionospheric storm time response. Driven by the initial dependence of the ground-based communications on HF radio, the negative storms were the first to receive the community's attention. Today, it is believed that the physical mechanisms of the negative storms are more or less understood, and negative ionospheric storm effects are mainly attributed to changes in neutral gas composition. In this scenario, the Joule and particle heating in the high-latitude region causes thermal expansion. In this way, large temperature and pressure gradients drive horizontal winds that produce global changes in the thermospheric circulation, with the storm-induced winds blowing from high to low latitudes. The divergence of the winds at subauroral latitudes causes the upwelling of neutral gas in the thermosphere, and by transporting neutral molecules into much higher altitudes, produces a neutral composition disturbance zone (also known as "composition bulge") and decreases the ionospheric ionization at F region heights [21,[24][25][26].
Neutral composition changes are present even during geomagnetically quiet conditions, but confined in the polar region. They are relatively small and restricted to subauroral latitudes, even during moderately disturbed geomagnetic conditions, especially in the daytime sector. Both the amplitude and the extent of the composition changes increase with increasing geomagnetic activity, and they are transported to lower latitudes by both the storm-induced and regular wind circulation [21,24].
Contrary to the case of negative ionospheric storms, positive storms have proven to be more challenging in their explanation. Even today, the research community has not arrived at a generally accepted explanation for their origin. Possible mechanisms proposed in the literature for the explanation of the observed positive storm effects mainly include [21][22][23]25,27,28]: i.
Neutral composition changes through the downwelling of neutral gas due to the convergence of neutral winds; ii. F layer uplifting due to vertical drifts. An increase in layer height leads to an increase in ionization density due to the reduced rate of ionization loss at higher altitudes. The F layer drift may be caused by two principal mechanisms: • Equatorward-directed winds: in this case, the ions and electrons feel a frictional force. The geomagnetic field-aligned component of this force pushes the ionization up to the inclined magnetic field lines, resulting in an uplifting of the F layer. Such winds may be caused either by large-scale wind circulation or as part of so-called traveling atmospheric disturbances (TADs). TADs are the superimposed result of impulse-like travelling disturbances formed by a wide spectrum of atmospheric gravity waves. These are launched by the injection of solar wind energy to the polar upper atmosphere. TADs propagate with high velocity towards lower latitudes carrying along equatorward-directed winds of moderate magnitude. • Electric fields: in this case, the height increase is caused by an ExB drift that is perpendicular to the inclined geomagnetic field, so it also leads to an uplifting of the ionosphere. Two major disturbance electric fields are: a. Disturbance dynamo electric fields: the disturbed thermospheric winds drive meridional winds to generate electric fields opposed to their quiet-time patterns. b. Penetration of electric fields originated from the solar wind/magnetosphere: high-latitude convection electric fields penetrate into the low-latitude ionosphere (a phenomenon known as prompt penetration electric fields, PPEFs) to produce changes known as the super-fountain effect [29].
iii. Advection of high-density plasma in combination to plasma uplift: nightside plasma is carried across the sunset terminator to draw out plumes of storm-enhanced density (SED) that stretches from the dusk sector to the noontime cusp ionosphere [30]. iv. Downward plasma flux from the plasmasphere.
Most probably, all these mechanisms may be considered important, with relevant contributions to the occurrence of positive ionospheric storms. However, vertical plasma drifts can also explain the observed increase in layer height that usually accompanies positive effects, and in this respect, they are claimed to be the dominant ones [21]. Among them, it has been suggested that thermodynamics may be more important for disturbances at the peak electron density height, while electrodynamics may more efficiently control the disturbances above [23] with the penetration fields, to play a significant role in the development of positive storms during the initial phase of the storm, in the afternoon sector, and at middle-to-low and low latitudes [22]. In any case, the relative importance of equatorward winds and electric fields in the generation of ionization increases, especially in the middle latitudes, remains among present challenges, and is not yet fully understood.

On the Occurrence of the Ionospheric Storm Effects
The occurrence of ionospheric storm effects is subject to several dependences for which the regular thermospheric winds play a key role. As an example, one may note the seasonal dependence: when storm-induced wind flow coincides with the direction of the prevailing meridional wind, which is the case for summer, the neutral composition disturbance zone penetrates to relatively low latitudes, and negative storms can also be observed at lower latitudes. On the other hand, in winter, the prevailing wind contrasts with storm-induced wind and, therefore, negative storm effects remain confined at higher latitudes to make positive effects the predominant feature of wintertime ionospheric storms [24][25][26]28].
Another key feature in the ionospheric storms occurrence is their dependence on local time. Neutral composition changes tend to be larger and extend to lower latitudes in the early morning sector. This is in agreement with the argument that the disturbance transport takes place in this sector due to the reduced ion drag and coincidence of the equatorward regular wind with storm-induced wind flow. In this respect, negative storm effects at middle latitudes are mainly observed during nighttime hours. While rotating into the daytime sector following the Earth's rotation, the disturbance zone is partially recovered and pushed back by poleward-directed daytime regular winds, but it is still able to produce daytime negative ionospheric storm effects. On the other hand, positive storm effects tend to occur in the daytime hemisphere, particularly in the prenoon and afternoon sectors [24,25,28]. Figure 2 provides an example of how positive and negative storm phases may alternate in different local time sectors and/or latitudes. For a more representative view of the ionospheric response to this intense storm event, the reader can refer to several research articles available in the literature, e.g., [31][32][33]. Multi-instrument data analysis and physicsbased model simulations are also carried out for other intense storm events that occurred in solar cycle 24, e.g., [34][35][36] provide the most recent evidence on the progress achieved in our present understanding, but also on the complexity that one may face in practice. Indeed, the sequence of ionospheric storm effects may be very complicated as several different processes work in concert, and the ionospheric response to each storm is determined by a unique and complex interaction of drivers, while multiple and/or successive intensifications of the geomagnetic activity may further complicate the interpretation of the observed effects, e.g., [37].
It has also been suggested that ionospheric storm effects are subject to solar cycle dependence. At the beginning of the 2000s, this argument was only based on model predictions. For instance, Burns et al. [38] found that, considering winter conditions, both the magnitude and extent of composition perturbations were larger during solar minimum, suggesting that disturbance effects are transported more efficiently out of the polar region into middle latitudes during these conditions. Interest in the solar cycle dependence of the ionospheric storm time response was greatly revived in the years following the deep 2008-2009 solar minimum, and probably the low activity solar cycle 24. Some indicative results came from identified shortcomings in ionospheric prediction capabilities during this deep solar minimum, under certain conditions that apply to all model formulations, from empirical to physics-based ones. For instance, Buresova et al. [39], in an investigation of the ionospheric disturbances under low solar activity conditions in 2008-2009, pointed out that the STORM model [40,41]-incorporated into the International Reference Ionosphere (IRI) standard model-did not provide reliable predictions for the peak electron density during CIR-generated storms. Wang et al. [42] also reported that the Coupled Magnetosphere Ionosphere Thermosphere (CMIT) model was not sufficiently able to reproduce the F2 peak electron density variations observed during the CIR events that occurred during the Whole Heliosphere Interval in 2008 (Day of Year 50-140). Such results may be considered to be suggestive of different ionospheric response to different levels of solar activity and/or different ionospheric response to geomagnetic storms driven by different solar/interplanetary conditions. The latter was also supported by the results obtained through the assessment of the Solar Wind-driven Autoregression Model for Ionospheric Forecast (SWIF) [43][44][45]. In particular, it was found that, while ionospheric disturbances related to the occurrence of CME-driven storms in the ascending phases of the solar cycles 23 and 24 were successfully predicted by the SWIF model, the performance of the model was significantly poorer during non-CME related ones which occurred in the same solar activity phase [46]. . The effects are more intense at lower latitudes (i.e., over Ebre). One may also note two qualitative differences in the response over the two distinct locations: (i) the positive storm penetrates the evening sector at lower latitudes, while at Chilton, it remains confined to daytime hours; (ii) Chilton experiences strong negative storm effects (foF2 is reduced by up to 50% with respect to the normal values) on the second storm day, most probably by entering the disturbance zone in the evening hours, while Ebre recovers to normal conditions. The work performed by Tsagouri et al. [47] supported an investigation of the longterm behavior of the ionospheric disturbances during the ascending phase of solar cycle 24 (i.e., for the time interval 2008-2014) with two objectives: the analysis of the ionospheric behavior in a wide range of geophysical conditions, and the determination of specific features that may differentiate the ionospheric response to different faces of space weather March. The effects are more intense at lower latitudes (i.e., over Ebre). One may also note two qualitative differences in the response over the two distinct locations: (i) the positive storm penetrates the evening sector at lower latitudes, while at Chilton, it remains confined to daytime hours; (ii) Chilton experiences strong negative storm effects (foF2 is reduced by up to 50% with respect to the normal values) on the second storm day, most probably by entering the disturbance zone in the evening hours, while Ebre recovers to normal conditions. The work performed by Tsagouri et al. [47] supported an investigation of the long-term behavior of the ionospheric disturbances during the ascending phase of solar cycle 24 (i.e., for the time interval 2008-2014) with two objectives: the analysis of the ionospheric behavior in a wide range of geophysical conditions, and the determination of specific features that Atmosphere 2022, 13, 346 8 of 13 may differentiate the ionospheric response to different faces of space weather driving. By extending a key part of their analysis here, Figure 3 provides the annual distribution of the DfoF2 as a metric of the ionospheric activity for the years 2008-2020. The results verify that significant ionospheric disturbances (greater than 20% with respect to normal conditions) are recorded during all years, even in extremely low solar activity conditions mainly capturing the impact of geo-effective solar wind structures. Furthermore, the results indicate that there are qualitative differences in the ionospheric activity between different levels of solar activity: the intensity of the positive disturbances tends to be higher during the years of low solar activity and solar minima (i.e., 2008-2010 and 2017-2019) compared to the solar maximum years (2012-2014), while they clearly dominate numerically over the negative disturbances. More generally, it was suggested that during low solar activity and solar minimum, the solar wind impact on the geospace environment particularly favors the occurrence of ionization increases, both in frequency and intensity. Ionization depletions are still present, but they mainly appear in moderate intensity, in contrast with the situation in solar maximum. The ionization increases were apparent in all local time sectors, while they clearly dominated in the afternoon hours, and they were accompanied only partially by an increase in the F2 layer peak electron density height (hmF2) [47]. Although it is known that the afternoon ionization enhancements observed during storm conditions, when portrayed as a percentage change, are inversely related to solar flux, simply because of the monthly median control curves having smaller magnitudes (i.e., net solar cycle effect) [22], the identified differences were better attributed to different solar wind drivers: positive disturbances with strong afternoon components tend to be a critical feature of the ionospheric response to the impact of CIR/HSSs in the ionospheric middle latitudes that tend to dominate during the declining solar cycle phase and solar minimum conditions. be higher during the years of low solar activity and solar minima (i.e., 2008-2010 and 2017-2019) compared to the solar maximum years (2012-2014), while they clearly dominate numerically over the negative disturbances. More generally, it was suggested that during low solar activity and solar minimum, the solar wind impact on the geospace environment particularly favors the occurrence of ionization increases, both in frequency and intensity. Ionization depletions are still present, but they mainly appear in moderate intensity, in contrast with the situation in solar maximum. The ionization increases were apparent in all local time sectors, while they clearly dominated in the afternoon hours, and they were accompanied only partially by an increase in the F2 layer peak electron density height (hmF2) [47]. Although it is known that the afternoon ionization enhancements observed during storm conditions, when portrayed as a percentage change, are inversely related to solar flux, simply because of the monthly median control curves having smaller magnitudes (i.e., net solar cycle effect) [22], the identified differences were better attributed to different solar wind drivers: positive disturbances with strong afternoon components tend to be a critical feature of the ionospheric response to the impact of CIR/HSSs in the ionospheric middle latitudes that tend to dominate during the declining solar cycle phase and solar minimum conditions. Similar results have also been reported at several occasions for middle latitudes. As indicative cases, we note: Buresova et al. [39] reported mainly positive disturbances in terms of the peak electron density under the occurrence of CIR-generated storms that occurred in the low solar activity years 2008-2009, with no significant changes in hmF2 in several cases; Burns et al. [48] found positive effects of geomagnetic storms in the peak electron density of the ionospheric F2 layer in the main storm days during CIR/HSSsdriven events; Dimitriev et al. [49] and Verkhoglyadova et al. [50,51] revealed the close relationship between the positive ionospheric storms and recurrent geomagnetic activity; Matamba and Habarulema [52] concluded that positive ionospheric storm effects in terms of GPS-derived TEC were more prevalent over middle, low and equatorial latitude stations during both CME-and CIR-driven storms, while negative ionospheric storm effects occurred only during CME-driven storms at the considered Southern and Northern Hemisphere middle latitudes.  Similar results have also been reported at several occasions for middle latitudes. As indicative cases, we note: Buresova et al. [39] reported mainly positive disturbances in terms of the peak electron density under the occurrence of CIR-generated storms that occurred in the low solar activity years 2008-2009, with no significant changes in hmF2 in several cases; Burns et al. [48] found positive effects of geomagnetic storms in the peak electron density of the ionospheric F2 layer in the main storm days during CIR/HSSs-driven events; Dimitriev et al. [49] and Verkhoglyadova et al. [50,51] revealed the close relationship between the positive ionospheric storms and recurrent geomagnetic activity; Matamba and Habarulema [52] concluded that positive ionospheric storm effects in terms of GPSderived TEC were more prevalent over middle, low and equatorial latitude stations during both CME-and CIR-driven storms, while negative ionospheric storm effects occurred only during CME-driven storms at the considered Southern and Northern Hemisphere middle latitudes.

Discussion and Conclusive Remarks
This paper attempts to provide a brief survey of knowledge on ionospheric storms, with the emphasis on the ionospheric effects at middle latitudes and the F region. As the ionosphere merges the influence from both above (solar wind/magnetosphere) and below (neutral atmosphere), ionospheric storms act as an important link in the complex chain of solar-terrestrial relations. In this respect, their investigation establishes a wide field of research that combines interdisciplinary expertise and/or knowledge. To keep the present survey brief, but at the same time, as complete as possible, the discussion is developed in a rather simplified way. For a fully justified overview of the present understanding on the ionospheric storms morphology and background mechanisms, the reader is urged to refer to extended reviews, collections of papers or chapter books available in the literature [1,[21][22][23][24][25]27,53,54].
Following many years of investigation, the physical mechanisms for the origin and occurrence of ionospheric storm effects are more or less in place. The situation is clearer in case of negative ionospheric storm effects, which are mainly attributed to changes in the neutral gas composition that are driven by Joule, and particle heating in the high-latitude region and consequent thermal expansion. Positive storm effects are more complicated in their explanation. They can be activated either also by neutral composition changes or the F layer uplifting, due to vertical drifts that bring the F layer higher where the rate of ionization loss is lower. The F layer drift may be caused by two principal mechanisms: equatorward-directed winds and electric fields (disturbance dynamo electric fields or the penetration of electric fields originated from the solar wind/magnetosphere). The advection of high-density plasma can take place in combination with plasma uplift, or through the interaction with the plasmasphere, and further explain observed ionization increases. Among the proposed mechanisms, vertical drifts seem to be the dominant ones, but the relative importance of equatorward winds and electric fields in the generation of ionization increases, especially in the middle latitudes, is still an open issue for the research community.
The occurrence of ionospheric storm effects is subject to several dependencies. Today, seasonal and local time dependencies are rather well justified and satisfactorily interpreted through the relative blow of storm-induced and regular thermospheric winds that keep the neutral composition disturbances confined to higher latitudes during daytime hours and the winter season. One of the key challenges in the investigation of the ionospheric storms in the last decade has been their solar cycle dependence. At the beginning of the 2000s, this dependence was rather an assumption based on model predictions [21,38]. However, interest in the solar cycle effect on the ionospheric storm time response was greatly revived in the years following the deep 2008-2009 solar minimum, and probably the low activity solar cycle 24. Shortcomings in ionospheric prediction capabilities during this deep solar minimum that were identified for all types of model formulations (from empirical to physics-based ones), the investigation of the long-term behavior of the ionospheric disturbances in solar cycle 24, and the analysis of specific storm events through multiinstrument data and models simulations, revealed the significant impact of the geomagnetic storm drivers in the observed effects [39,42,[46][47][48][49][50][51][52]. Although it is known that ionospheric storm effects of significant intensity are observed even during solar minimum years, there is now some evidence that the ionospheric response to CIR/HSSs-driven storms may be different than the one to CME-driven events. In particular, it has been suggested that positive disturbances with strong afternoon components tend to be critical features of the ionospheric response to the impact of CIR/HSSs in the ionospheric middle latitudes, while these dominate during the declining phase and solar minimum conditions [47].
At first approach, the differences in the ionospheric storm-time response may be attributed to the intensity of the storm events: as CIR/HSSs drive geomagnetic storms of weak to moderate intensity, one would expect the generated neutral composition disturbance zone to be confined to higher latitudes. In any case, it may also be expected that the relative contribution of the ionospheric storm background mechanisms may differ between CME-and CIR/HSSs-driven disturbances, as the result of the differences in the total energy deposition into the auroral atmosphere [55][56][57], and the way in which this energy is distributed [15,58] during different storm scenarios. Of particular importance may be the role of the particle precipitation in the development of the ionospheric response during CIR/HSSs-driven disturbances [47]. The resulting heating may cause large-scale changes to the thermosphere composition, density and circulation (winds) in any case [21,59], but the rate of the energy deposition may differ between CME and CIR/HSSs cases [60]. It has been reported that the main phase of CME-driven events is characterized by intense but short-lived particle precipitation excursions, while the non-CME cases are characterized by prolonged periods of particle precipitation [47], which in turn may result in the gradual development of the disturbances and the limited extent of the neutral composition disturbance zone. In any case, the clarification of how the upper atmosphere is affected by solar coronal holes still requires the attention of the research community. Present availability in global-scale observations, new models and new analysis methods may facilitate systematic studies. The results, in combination with the output of studies on the CMEs that dominate solar maximum, would complete our understanding of storm time ionospheric response throughout the solar cycle [61], and would significantly advance our prediction capabilities in case it is proven that different storm scenarios need to be addressed differently by ionospheric prediction models [47,62,63].
A final point for further consideration when discussing ionospheric storm effects, is that it is not yet fully clarified whether foF2/NmF2 and TEC storm time responses always share the same features, and therefore, results obtained from the analysis of data from different data sources should be considered cautiously. TEC is defined as the integral of the entire ionospheric electron density profile. Such an integral has its maximum contribution from the F2 layer, with approximately 2/3 of the TEC coming from regions above the altitude (hmF2) of peak density (NmF2) [64]. Therefore, if substantial effects occur in TEC, simple vertical redistributions of F layer plasma cannot be the main cause [22]. In addition, in case the electron concentration at heights of the layer maximum is more sensitive to changes in thermospheric processes, and the effects of electrodynamics influence strongly higher altitudes which contribute substantially to the TEC estimates, one may expect quantitatively and maybe qualitatively different responses between foF2/NmF2 and TEC. Indeed, such discrepancies have been reported in the literature, such as, for instance, the results from the analysis of Chen et al. [65]. The authors conducted an epoch analysis of global ionosphere responses to recurrent geomagnetic activity during 79 corotating interaction region (CIR) events from 2004 to 2009 using GPS-derived TEC values, to find that in the high and middle latitudes, the TEC data showed a positive effect on the first storm day, while the negative response in the following days was only evident in the high latitudes. However, the daytime TEC response to CIR storms appeared to be different from that of ionospheric F2 peak densities at high and middle latitudes, as NmF2 had a long period of negative response for about 2-3 days, whereas the TEC response was mostly positive. Liu et al. [66], by conducting an analysis of the storm event of March 2015 at middle latitudes, reported SED signatures in TEC, but not in NmF2. Another example comes from the SWIF model scenarios: the comparative analysis of the variations in foF2 and TEC during eleven geomagnetic storm events that occurred in solar cycle 24 revealed similarities but also differences in the storm-time response of the two characteristics with respect to the local time and the latitude of the observation point. The most important difference concerns the intensity of positive storm effects, which appeared significantly greater in the TEC case. Moreover, positive storm effects at higher latitudes in case the observation point at storm onset time is located in the morning sector were present only for TEC. The TEC positive phase was more intense in the lower latitudes, especially in the afternoon hours [67].

Conflicts of Interest:
The author declares no conflict of interest.